Skip to main content

LinkedIn's Stream Experimentation Framework

Joseph Adler (Facebook), Xin Fu (LinkedIn Corporation), Bee-Chung Chen (LinkedIn, Inc.)
Hadoop and Beyond
GA Ballroom J
Average rating: ****.
(4.00, 17 ratings)

At LinkedIn, we’re constantly looking for ways to deliver more insights and recommendations to our users. The only way to learn what users find useful is to test things, so we constantly work on making tests cheaper and more effective.

In this talk, we’ll review our experimentation framework for the home page. We’ll talk about the three key systems that make this possible:

- The framework for experiment design, deployment, and analysis. (Xin Fu)
- The system for adding content to the stream. (Joseph Adler)
- The ranking system for finding and prioritizing the best content. (Bee-Jung Chen)

We’ll provide some real examples of how we use this system, and talk about some of the lessons that we’ve learned.

Photo of Joseph Adler

Joseph Adler

Data Science, Facebook

Joseph Adler has many years of experience in data mining and data analysis at companies including DoubleClick, American Express, and VeriSign. He graduated from MIT with an B.Sc. and M.Eng in Computer Science and Electrical Engineering. He is the inventor of several patents for computer security and cryptography, and the author of “Baseball Hacks” and “R in a Nutshell”. Currently, he is a senior data scientist at LinkedIn.

Xin Fu

Principal Data Scientist Lead, LinkedIn Corporation

To be added later

Bee-Chung Chen

Staff Engineer, LinkedIn, Inc.

To be added later